Decoupling feature extraction from policy learning: assessing benefits of state representation learning in goal based robotics.
Antonin RaffinAshley HillKalifou René TraoréTimothée LesortNatalia Díaz RodríguezDavid FilliatPublished in: CoRR (2019)
Keyphrases
- feature extraction
- reinforcement learning
- learning algorithm
- learning process
- learning systems
- online learning
- learning scheme
- training data
- prior knowledge
- supervised learning
- knowledge acquisition
- action selection
- neural network
- learning scenarios
- learning tasks
- collaborative learning
- state space
- active learning
- decision trees
- image processing